The purpose of this study is to test the hypothesis that human milk beta-glucuronidase is a significant factor related to neonatal hyperbilirubinemia. Bilirubin metabolism will be compared in three groups of newborn human infants. The diet of the three groups will differ in beta-glucuronidase content: high beta-glucuronidase (breast milk), low beta-glucuronidase (standard formula), and beta-glucuronidase inhibitor (casein hydrolysate).